Computing and Information Systems - Theses
Permanent URI for this collection
Search Results
1 - 6 of 6
-
ItemLazy Constraint Generation and Tractable Approximations for Large Scale Planning Problems( 2023-12)In our research, we explore two orthogonal but related methodologies of solving planning instances: planning algorithms based on direct but lazy, incremental heuristic search over transition systems and planning as satisfiability. We address numerous challenges associated with solving large planning instances within practical time and memory constraints. This is particularly relevant when solving real-world problems, which often have numeric domains and resources and, therefore, have a large ground representation of the planning instance. Our first contribution is an approximate novelty search, which introduces two novel methods. The first approximates novelty via sampling and Bloom filters, and the other approximates the best-first search using an adaptive policy that decides whether to forgo the expansion of nodes in the open list. For our second work, we present an encoding of the partial order causal link (POCL) formulation of the temporal planning problems into a CP model that handles the instances with required concurrency, which cannot be solved using sequential planners. Our third significant contribution is on lifted sequential planning with lazy constraint generation, which scales very well on large instances with numeric domains and resources. Lastly, we propose a novel way of using novelty approximation as a polynomial reachability propagator, which we use to train the activity heuristics used by the CP solvers.
-
ItemWorkflow Scheduling in Cloud and Edge Computing Environments with Deep Reinforcement Learning( 2023-08)Cloud computing has firmly established itself as a mandatory platform for delivering computing services over the internet in an efficient manner. More recently, novel computing paradigms such as edge computing have also emerged to complement the traditional cloud computing paradigm. Owing to the multitude of benefits offered by cloud and edge computing environments, these platforms are increasingly used for the execution of workflows. The problem of scheduling workflows in a distributed system is NP-Hard in the general case. Scheduling workflows across highly dynamic cloud and edge computing environments is even more complex due to inherent challenges associated with these environments including the need to satisfy diverse contradictory objectives, coordinating executions across highly distributed infrastructures and dynamicity of the operating conditions. These requirements collectively give rise to the need for adaptive workflow scheduling algorithms that are capable of satisfying diverse optimization goals amid highly dynamic conditions. Deep Reinforcement Learning (DRL) has emerged as a promising paradigm for dealing with highly dynamic and complex problems due to the ability of DRL agents to learn to operate in stochastic environments. Despite the benefits of DRL, there are multiple challenges associated with the application of DRL techniques including multi-objectivity, curse of dimensionality, partial observability and multi-agent coordination. In this thesis, we propose novel DRL algorithms and architectures to efficiently overcome these challenges.
-
ItemCapturing Uncertainty in Ensemble Models For Human-Machine Collaboration( 2023-08)This thesis studies capturing uncertainty in ensemble models, starting with machine only models and then leading to human-machine collaboration from new perspectives. We have identified two research gaps in the previous studies on capturing uncertainty in ensemble models. First, in machine decision making, while several studies have been presented to introduce combining approaches to deal with inter-source uncertainty in ensemble models, capturing intra-source uncertainty needs to be addressed. Second, the collaboration of humans with machines in ensemble models introduces a new challenge of integrating human uncertainty within these models. Furthermore, when addressing real-world decision making problems through human-machine collaboration, dedicated research efforts are required to investigate this collaborative approach thoroughly. So, each case study in human-machine collaboration for ensemble models requires new approaches. This thesis delves into the first research gap in the context of ensemble classifiers. It studies this research gap in two settings. First, an interval modelling to combine classifiers in a category of ensemble models to capture inter-source and intra-source uncertainties is developed. Through the proposed model, the performance of the ensemble model can be improved by capturing uncertainty in complicated binary classification problems. Second, the ensemble selection problem for bagging as one of the widely used ensemble models in the literature is studied. While consideration of intra-source uncertainty as a selection criterion for classifiers in an ensemble has been previously overlooked, we show its substantial to enhance the performance of the selected ensemble. This study formulates the ensemble selection problem as a bi-objective optimisation problem for bagging and presents an adaptive meta-heuristic algorithm to solve the bi-objective problem. The findings highlight the significance of incorporating intra-source uncertainty into the classifier selection process, leading to improved ensemble model performance. Then we touch upon the second research gap on human-machine collaboration in ensemble models. Specifically, interval modelling approaches to capture uncertainty in ensemble models where both humans and machines make decisions are presented. Through two case studies, we show how this collaboration and interval modelling enhance ensemble performance, by capturing uncertainties from both humans and machines. In the first case study, synthetic data is used to show how human-machine collaboration and capturing the uncertainty of humans and machines can be conducted throughout the ensemble decision making. In the second case study, a real image dataset related to biofouling assessment is utilised to investigate the importance of human-machine collaboration in ensemble models for biofouling detection. Furthermore, interval modelling to examine how capturing uncertainty affects ensemble performance is employed for biofouling detection. In summary, as we explained, this thesis advances the current state of ensemble models by presenting effective interval modelling techniques to capture uncertainty and investigates human-machine collaboration within these models. These contributions aim to enhance ensemble performance, a critical step before implementing these models to deal with real-world decision making problems.
-
ItemExplainable Reinforcement Learning Through a Causal Lens( 2021)This thesis investigates methods for explaining and understanding how and why reinforcement learning agents select actions, from a causal perspective. Understanding the behaviours, decisions and actions exhibited by artificially intelligent agents has been a central theme of interest since the inception of agent research. As systems grow in complexity, the agents' underlying reasoning mechanisms can become opaque and the intelligibility towards humans can be diminished, which can have negative consequences in high-stakes and highly-collaborative domains. The explainable agency of an autonomous agent can aid in transferring the knowledge of this reasoning process to the user to improve intelligibility. If we are to build effective explainable agency, a careful inspection of how humans generate, select and communicate explanations is needed. Explaining the behaviour and actions of sequential decision making reinforcement learning (RL) agents introduces challenges such as handling long-term goals and rewards, in contrast to one-shot explanations in which the attention of explainability literature has largely focused. Taking inspirations from cognitive science and philosophy literature on the nature of explanation, this thesis presents a novel explainable model ---action influence models--- that can generate causal explanations for reinforcement learning agents. A human-centred approach is followed to extend action influence models to handle distal explanations of actions, i.e. explanations that present future causal dependencies. To facilitate an end-to-end explainable agency, an action influence discovery algorithm is proposed to learn the structure of the causal relationships from the RL agent's interactions. Further, a dialogue model is also introduced, that can instantiate the interactions of an explanation dialogue. The original work presented in this thesis reveals how a causal and human-centred approach can bring forth a strong explainable agency in RL agents.
-
ItemMulti-Granular Webpage Information Extraction and Analysis via Deep Joint Learning( 2020)The number of webpages is growing exponentially, which results in a great volume of unstructured information on the web. It takes time either to fully comprehend a webpage or to retrieve relevant information from a complex webpage. Analyzing unstructured webpage and extracting structured information from the webpage automatically is crucial. In this study, we aim to develop algorithms for multi-granular webpage information extraction and analysis to facilitate webpage information understanding. We investigate the problem at three levels of granularity, i.e., micro, meso and macro levels. For every level, we focus on one extraction and analysis task, although the algorithms we developed are general and can be applied to many other similar tasks. At the micro level, we aim to extract webpage entities that have diverse forms, and focus on the application of person name recognition. We propose a fine-grained annotation scheme based on anthroponymy and create the first dataset for fine-grained name recognition. We propose a joint model that learns the different name form classes with two sub-neural networks while fusing the learned signals through co-attention and gated fusion mechanisms. Experimental results show that our annotations can be utilised in different ways to improve the recognition performance. At the meso level, we study the relationships between webpage entities and blocks with a focus on the application of joint recognition of names and publications. We address the person name recognition and publication string recognition tasks in academic homepages jointly based on the insight that the two tasks are inherently correlated. We propose a joint model to capture the interdependencies between entities. We also capture global position patterns of blocks and local position patterns of entities in the model learning process. Empirical results on real datasets show that our model outperforms the state-of-the-art publication string recognition model and person name recognition model. Experimental results also show that our model outperforms baseline joint models. At the macro level, we aim to provide hierarchical analysis for webpages from diverse domains. We introduce the Webpage Briefing (WB) task, which aims to generate a summary of a webpage in a hierarchical manner, starting at the top is an abstract and general description of the topic of the webpage page, followed by high level key attributes extracted from the webpage, and then lower level key attributes, which contain concrete and specific key information. We propose to perform webpage briefing by identifying and summarizing the informative contents, which mimic human behaviour of understanding a complex webpage. We propose a novel Dual Distillation method that has a teacher-student architecture with dual distillation. We further propose a Triple Distillation method to better exploit the inherent correlation of specific key attributes and general topics of webpages. We finally propose a novel Triple Joint model that has a triple joint learning architecture with signal exchange and enhancement mechanisms. Experimental results show the superiority of Bi-Distill method and Tri-Distill over baseline methods. Experimental results also show that Tri-Join outperforms baseline single-task models and baseline jointly trained models.
-
ItemCost-efficient Management of Cloud Resources for Big Data Applications( 2020)Analyzing a vast amount of business and user data on big data analytics frameworks is becoming a common practice in organizations to get a competitive advantage. These frameworks are usually deployed in a computing cluster to meet the analytics demands in every major domain, including business, government, financial markets, and health care. However, buying and maintaining a massive amount of on-premise resources is costly and difficult, especially for start-ups and small business organizations. Cloud computing provides infrastructure, platform, and software systems for storing and processing data. Thus, Cloud resources can be utilized to set up a cluster with a required big data processing framework. However, several challenges need to be addressed for Cloud-based big data processing which includes: deciding how much Cloud resources are needed for each application, how to maximize the utilization of these resources to improve applications' performance, and how to reduce the monetary cost of resource usages. In this thesis, we focus on a user-centric view, where a user can be either an individual or a small/medium business organization who want to deploy a big data processing framework on the Cloud. We explore how resource management techniques can be tailored to various user-demands such as performance improvement, and deadline guarantee for the applications; all while reducing the monetary cost of using the cluster. In particular, we propose efficient resource allocation and scheduling mechanisms for Cloud-deployed Apache Spark clusters.